What’s Confusion Matrix
Confusion Matrix, is a matrix and… we all agree.
Confusion Matrix can answer to the questions:
- How is going the prediction?
- Which one I constantly missed?
Confusion Matrix is composed by following components:
What’s TP?
- True Positive: it’s what the model have predicted right. On my current work I’m trying to predict the weight of a box (about
11classes). The box weight750? Yes, it falls into TP recipient. If not, it falls into FP.
What’s FP?
- False Positive: it’s what the model have predicted wrong. The model thought the box weight
750KG but it isn’t. Wrong prediction.
What’s TN:
- True Negative: the model understood that box weight isn’t
750KG. It falls into TN recipient. If not, it falls into FN.
What’s FN:
- False Negative: the model predicted
not 750KG and was wrong. It falls into FN recipient. it’s the opposite behavior of FP. Someone would say: invert. always invert
When Confusion Matrix
This kind of matrix is applicable only on Classification Problem.
Why Confusion Matrix
On top of Confusion Matrix have been developer lots of metrics. :
Implementation
If you have any suggestions, recommendations, or corrections please reach out to me.